The low rate of women employed in the construction sector in Spain is not a phenomenon exclusive to this country (13), the construction industry is the most male-dominated sector in all countries of the world (14), (15). Percentages regarding the labor market access are similar in different countries. These figures have only slightly changed in relation to the last years (16)-(18). According to the International Labor Organization (ILO), in 2018, the percentage of women employed in the construction sector reached 10% in the case of the United States; in India, for that same year, the percentage was 8%, and Argentina did not reach 4% (19).

Women’s unequal access to the labor market in the construction sector is increasingly under study in scientific literature in different countries (8), (10), (11), (17), (20-30). These investigations share the vision of the construction sector as one of the labor markets being more associated with gender stereotypes. The construction sector is believed to be a rough and dirty sector, an inappropriate place for femininity (10), (18), (31)-(33). In a highly male-dominated working environment, women are often left out. This hinders their entrance and participation in the building labor market (10), (20), (21), (34)-(36). The limited opportunities to enter the world of work entail low labor market participation. As a consequence, there is a lack of work experience, and less accumulated experience implies difficulties in finding a job (37). Therefore, promoting women’s access to the construction sector is a critical issue.

Apart from the obstacles faced by women that hamper their entrance to the construction industry, there are other barriers met by female employees in this field throughout their professional careers (36), (38). These barriers hinder the exercise of their work, the fact of obtaining positions of power, and finally, their continuity in the sector (8), (28), (29). Among the most studied barriers, we could mention the incompatibility of the particular features of the sector (such as the lack of flexible working hours), the nomadic nature of site works, presenteeism, or long working hours together with domestic, and family responsibilities (8), (10), (20), (28), (36). Another reported obstacle is the discrimination suffered by women in the recruitment or promotion procedures in the construction sector. On the one hand, it is due to the existence of stereotypes, as still today, men are expected to be better in project management than women. On the other hand, there are informal male exclusive contact networks, where contracts and promotions are negotiated, among other issues. This is known as a homosocial behavior phenomenon (8), (10), (18), (20), (28), (34), (36)-(42).

In the light of the foregoing, and for the commitment with gender equality and achievement of the Sustainable Development Goals (43), the Spanish General Council of Technical Arquitecture (CGATE, in Spanish) has conducted a survey among professionals working in the sector. The survey covered the situation of women in the construction industry (44), (45) and the goal of this project is to analyze the findings revealed in that survey. Thus, we could know the attitude of Spanish building engineers towards gender equality, and determine if their attitude and perception towards equality differs according to the gender of professionals in this sector. This way, we could draw conclusions supported by a bivariate analysis that will enable us to know the real situation on equality faced by building engineers in Spain in the exercise of their profession.

Participants of this study are a representative sample of the professionals belonging to associations of Official Colleges of Building Engineers in Spain in 2018. At that time, there was a total of 49821 members, 79.2% were men (39302 professionals), and 20.8% female (10341 professionals) (46). Among the 1353 interviews that were conducted, 659 respondents were male (48.7%), and 694 were female (51.3%), see Table 1. Assuming simple random sampling, and in the most unfavorable hypothesis being p=q=0.5, the sampling error for the total sampling is ±2.7%. For men, the sampling error is ±3.9, and for women it is ±3.8%. The questionnaire was administered by Computer Assisted Web Interviews (CAWI), using Google Forms. The fieldwork was done between September and December 2018, and the Spanish General Council of Technical Arquitecture was in charge of this task. Once the data are collected, the Spanish General Council of Technical Arquitecture gave the right of dataset exploitation to the current research team for analysis.

To ensure the ethics of scientific research, the design of the questionnaire and its implementation was supervised by the Data Protection Officer of the Spanish General Council of Technical Arquitecture. The link to the online survey was provided through a banner in the same webpage of the Spanish General Council of Technical Arquitecture. In order to gain access to the questions, participants had to previously read a document where the goals and objectives of the study were presented. There were also issues such as the approximate duration of the questionnaire and the willingness and anonymity guaranteed by the Data Protection Law existing in Spain (47).

The questionnaire was designed by the Spanish General Council of Technical Arquitecture. It is composed of 39 close questions, being four of them sociodemographic questions which will allow us to characterize the sample (see Table 1). The rest of questions measure the opinion of workers towards different issues related to gender equality, reconciliation of family life, equal pay or equal treatment in the Building Engineering sector.

The bulk of the sample (more than 60%) was aged between 30 and 50 years old, 9.6% being under 30 years of age, and 6.4% over 60 years of age. As regards how the profession is done, 59.5% carry on their profession on a self-employed basis, followed by 28.9% who work as an employee for others. As far as monthly revenues are concerned, more than 50% of the sample earn between 1000 and 2000 €, although it should be noted that, over a quarter of the sample (20.6%) are paid less than 1000 € per month. When disaggregating data according to gender and age, for ages above 51, there is a greater presence of men (32.9%) than women (10.1%). Furthermore, there are more women than men working as an employee for others, or in the public sector. In relation to the income, the results indicate a higher presence of women in the lowest areas. 18.3% of women earn between 500 and 1000 €, as opposed to 12.5% of men who are in the same salary level. On the contrary, those receiving a larger salary are mainly men.

The questions raised in the analysis are linked to the respondents’ perception of gender equality. Firstly, they were asked about their perception towards gender equality in general terms; secondly, about equality in the personal working environment, and thirdly, about a future readiness to promote equality in the sector (Table 2).

In order to know the respondents’ answers and the sociodemographic characteristics associated to the answers, a bivariate analysis has been conducted using the Chi-square test. Furthermore, the strength association is determined with Cramer’s V or φ (Phi), according to nxm or 2x2 contingency tables (48). Both statistics (Cramer’s V and φ) indicate the strength of association with a value between 0 and 1; the closer it is to 1, the greater the strength of association of the variables analysed. When dealing with variables of a social nature, a high strength of association is considered from values at 0.3 (48). Apart from that, we thoroughly studied the categories where a greater association between the sociodemographic and the content variables is displayed; that is, going beyond the variable analysis to go into much more detail with the study of the corrected typified wastes (48). In the contingency tables rendered in the next section, it will be indicated with arrows, those cells whose corrected typified wastes (the typified difference between the observed and expected frequency) are higher than 1.96. This means that the association between categories has a reliability level higher than 95.5%. The following is the notational system used to represent this: up arrows (↑) representing a positive association, that is, when there are more cases than expected in a random distribution; and down arrows (↓) when there are fewer cases than expected in a random distribution, and therefore, that entails a negative association.

One of the first questions addressed to the sample was the perception of the gender who encounters more difficulties in entering the building engineering profession, men, women, or both genders equally. As is evident from Table 3, 59.1% of the sample (800 cases) believes that it is more difficult for women to enter the profession. 40.1% (542 cases) considers both genders, men, and women equally, and 0.8% (11 cases) admits that men will be affected the most when entering the profession.

In judging these results according to those sociodemographic characteristics presenting a statistical association, the gender of the respondent shows a stronger association (one of the highest in the study, Cramer’s V = 0.330). In fact, there are wide percentage differences in the replies given by men and women. Women consider, to a greater extent, that women encounter more difficulties in entering the profession. 74.9% of women certainly think that, as opposed to 42.5% of men who hold the same vision. There are large statistically meaningful percentage differences in regards to the opinion of both genders as having problems when entering the profession. 56.6% of men take this view, as opposed to 24.5% of women. Accordingly, whereas most women claim that women are the worst affected gender when entering the profession, the vast majority of men do not see any differences. Men believe that both genders are likewise affected.

Another statistically significant association of this content variable with the variable ‘age’ (Cramer’s V = 0.108) and the variable ‘exercise of the profession’ (Cramer’s V = 0.108) is also present. When analyzing the corrected typified wastes, younger age groups (below the age of 30) consider to a greater extent that the gender who encounters more difficulties when entering the profession is the female gender. They account for 73.8% (96 cases). The older respondents (those that are 51-60 and above) believe, to a greater extent, that no differences between genders are found when entering the profession.

Finally, with regards to the ‘exercise of the profession’, the most critical individuals with the situation of women are the salaried employees (69.6%, 266 cases). However, the self-employed individuals consider, to a lesser degree, that no differences are seen regarding gender when entering the profession (46.0%, 370 cases).

In the survey, another question addressed to the respondents made them reflect on the gender that a client or developer would choose when hiring a technical service. In Table 4, we can observe how 54.7% (740 cases) of the sample considers that the client or developer would hire equally, and regardless of gender. 43.6% (590) indicates that the client or developer would hire a man. Finally, 1.6% (22 cases) declares that the client or developer would prefer a woman for this technical service.

Once more, the variable ‘sex’ of the respondent discloses a high statistical association (Cramer’s V = 0.285). If attention is paid to the corrected typified wastes, men believe, to a great extent (68.9% - 454 cases), that the developer or client would hire either gender. Nonetheless, women who have participated in the survey think differently. Most women (57.4%, 398 cases) admit that the developer or client would rather hire a man for the technical service.

Although the strength of association is not that high, the variables ‘exercise of the profession’ (Cramer’s V =0.094) and ‘monthly income’ (Cramer’s V = 0.098) also present a significant statistical association with this question. Regarding the ‘exercise of the profession’, salaried employees (51.0%, 195 cases), followed by officials or similar (50.9%, 84 cases) perceive a higher inequality. They consider that men would be the preferred choice of the developer or client. Meanwhile, self-employed individuals (60.1% of this specific group, 484 cases) consider to a great extent, that the client or developer will not take into account the technician’s sex. Finally, those having the highest salary (more than 3000 € per month) believe to a much greater extent that the developer or client would hire a person for the technical service, without regard to the gender (72.0%, 59 cases).

In another question of the survey, respondents were asked about the ease to be hired in a company of the sector. On this occasion, as displayed in Table 5, 61.9% of the sample (837 cases) claims that it is easier for men to be hired. 35.4% (479 cases) does not find any differences in relation to gender when being hired by a company. 2.7% (36 cases) consider that it is easier for women to be hired.

As displayed in Table 5, other sociodemographic variables that are statistically associated with this question are ‘age’ (Cramer’s V= 0.117), ‘exercise of the profession’ (Cramer’s V= 0.086), and ‘monthly income’ (Cramer’s V= 0.100). With respect to the variable ‘age’, the corrected typified wastes reveal that the percentage differences in the older groups present a statistical association. The respondents pertaining to the age groups 51-60 (52.2%, 105 cases) and above 60 (48.8%, 42 cases) consider to a lesser extent that it is easier for men. Respondents in the age group 51-60 consider to a great extent that there are no gender-based differences when being selected by a company, as this process is not linked to gender (44.8% of this group, 90 cases).

The variable ‘exercise of the profession’ preserves the dynamics of the questions discussed above. The salaried employees are the most critical respondents by reporting to a great extent that it is easier for men (68.8%, 263 cases). Nevertheless, the self-employed group claims that the staff selection does not depend on gender (39.6%, 319 cases), as the corrected typified wastes study show. Finally, regarding the ‘monthly income’, those receiving higher salaries (more than 3000 €) consider to a great extent (57.3%, 47 cases) that the staff selection does not depend on gender. This makes it to be statistically significant.

In order to finalize the analysis of questions related to the perception of equality in the Technical Architecture sector, the opinion concerning equal pay for men and women has been analyzed in Table 6. 54.4% of the sample set (735 cases) believes that there is still a wide pay gap between men and women, whereas 45.6% of the sample declares that there are no disparities (616 cases).

As displayed in Table 6, the Chi-square test study brings back statistical significance in all the sociodemographic variables outlined in the study. In the variable ‘sex’, the strength of this association is the highest association throughout the inquiry (φ = 0.369). The study of the corrected typified wastes in this variable reveals that men consider to a great extent that wages are equal (64.4%, 424 cases), whereas a vast majority of women admitted not to be remunerated in the same fair manner as their male colleagues (72.3%, 501 cases).

In regards to the variable ‘age’ (Cramer’s V = 0.158), the same dynamic can be seen. Older age groups perceive less inequality, whereas the youngest groups are the most critical. This time, respondents between 30 and 40 years old believe to a great extent that there is a discrepancy between the wages levels paid between male and female workers, to the detriment of women (60.4%, 249 cases). People in the 51-60 age group (57.2%, 115 cases) together with the group above 60 (66.3%, 57 cases) claim that wages are indeed equal. Regarding the variable ‘exercise of the profession’, the salaried employees mainly report the gender inequality in wages. 67.5% (257 cases) of salaried employees believe that women wages are not fair vis-à-vis their male colleagues. Nevertheless, respondents who declare working as self-employed believe that wages are equal (51.4%, 414 cases).

Finally, when examining the variable ‘monthly income’ we can once more observe that those having higher salaries (more than 3000 € per month) are the least critical with the pay gap. 61.0% (50 cases) believe in the existence of a fair wage gap between men and women. Nonetheless, respondents earning between 1001 and 1500 € per month are the most critical age group in this question. 59.7% (237 cases) claim that salaries are not distributed equally amongst men and women in the Building Engineering sector, being women the worst affected group.

In this section, we are going to analyze two questions related to the equality in their own company, or in their job. Firstly, respondents are requested to indicate the gender of the majority of their colleagues. As can be seen in Table 7, 68.9% of the sample believes that their colleagues are mainly male (931 cases), followed by 23.2% (314 cases) who believe that there is an equal percentage of men and women. 7.9% (107 cases) of the sample declares that their colleagues are mainly women. In this occasion, the bivariate analysis of this question conducted through the Chi-square test brings back statistical association with the following sociodemographic variables: ‘sex’ (φ = 0.326) and ‘exercise of the profession’ (Cramer’s V = 0.118).

As far as the variable ‘exercise of the profession’ is concerned, we should outline the percentage of self-employed individuals (71.8%, 578 cases) who believe that their colleagues are mainly men. This situation changes in officials or similar, as 31.5% of them declare that their colleagues are men and women, in the same proportion (52 cases). It is worth stressing that, it is within this group where we find the highest percentage of respondents who esteem that their colleagues are mainly women, reaching 17.6% (29 cases).

A second question related to the personal experience in the working environment requested respondents to consider whether or not, in their opinion, in their working environment, opinions coming from men and women are equally valued. In accordance with Table 8, a majority agreement is reached in the sample set. 76.0% (1028 cases) admit that opinions coming from men and women are equally valued, regardless of the gender of the person. 24.0% (324 cases) has a different view of the matter.

The distribution of answers is statistical significant in all the variables included in the analysis (Table 8). In this sense, the variable ‘sex’ presents again a greater strength of association (φ = 0.326), followed by the variables ‘monthly income’ (Cramer’s V = 0.124), ‘age’ (Cramer’s V = 0.116) and ‘exercise of the profession’ (Cramer’s V = 0.115). Focusing attention on the corrected typified wastes, significant percentage differences in men’s and women’s opinion are observed, being women the group who believes the most that opinions coming from men and women are not equally valued (37.5%, 260 cases). The 90.3% (593 cases) of men participating in the study claims that opinions coming from men and women are equally valued.

Concerning the ‘exercise of the profession’, the most critical group with the fact of valuing the opinion of men and women alike are the salaried employees, where 30.9% (118 cases) do not think that it is valued alike. On the contrary, self-employed agree the most with the fact of opinions being equally valued (80.0%, 644 cases). Finally, after due analysis of the variable ‘monthly income’, it appears that individuals earning between 1001 and 1500 € are those who believe the most that opinions are not equally valued, whereas individuals with higher wages (2001-3000 € and more than 3000 €) consider to a great extent that opinions are equally valued.

In order to conclude the results section, we will analyze the respondents’ attitude or readiness to promote equality in the Building Engineering sector. In the first question, shown in Table 9, respondents were asked about the possibility to take actions that improve the reconciliation of family life and work. There was a broadly shared view in the sample set, that is, 93.1% (1259 cases) believe in the possibility to take actions that improve the reconciliation of family life and work. 6.9% (93 cases) do not share this view.

The sociodemographic variables that expose a statistical association with this question are ‘sex’ (φ = 0.092) and ‘exercise of the profession’ (Cramer’s V = 0.075); both with a low strength of association. The study about the corrected typified wastes found that again women consider the most that actions improving the reconciliation could be conducted (95.4%, 661 cases). A view that is shared by a vast majority of men, although the percentage is lower, as it dropped significantly, 79.5% (524 cases). Finally, differences in ‘exercise of the profession’ reveal a high level of agreement regarding the actions to be carried out in order to improve the reconciliation in participants working as salaried employees (95.5%, 365 cases).

In the last question, respondents were asked whether or not actions should be taken by the professional organization in order to study the real situation of women in the Building Engineering sector. By comparison with the previous question, Table 10 indicates that although the percentage of those in agreement with this measure is high, it decreases to 86.5% (1169 cases). With reference to the statistical association of this question with the sociodemographic variables included in the analysis, ‘sex’ (φ = 0.198) and ‘exercise of the profession’ (Cramer’s V = 0.143) are the variables that bring back a significant statistical.

The analysis reveals that men manifest more their opposition to the actions taken by the professional organization in order to study the real situation of women in the construction sector. This opposition is declared by 20.5% of the male respondents (183 cases). In the case of women, 93.1% (645 cases) consider that actions should be taken by the professional organization. Finally, self-employed individuals show again a great rejection to this kind of actions. 17.5% (141 cases) believe that this kind of actions should not be carried out.